Fabric IQ: The shared context layer for AI agents and real-time applications AI agents fail without context-rich data. A June 2026 Forrester report puts it plainly: "No matter how well modeled, data without the context provided by shared semantics and mature ontologies will fall short for agentic AI use cases." Microsoft's answer at Build 2026 is Fabric IQ, now generally available as the shared context layer sitting at the heart of the Microsoft data stack. The architecture is what makes this interesting from a graph and semantics perspective. Fabric IQ is built on three connected layers: Unified data in OneLake (structured, unstructured, event, and graph), business intelligence through semantic models (business meaning defined once, applied everywhere), and operational intelligence through ontologies and real-time signals. Ontologies are doing the heavy lifting here. They model the entities, relationships, rules, and constraints that semantic models were never designed to represent.  According to a Microsoft-controlled A/B study, ontology-grounded agents produced 2.2x more excellent responses, a 4.5x win rate in side-by-side comparisons, and a 30% reduction in tool calls. Without ontology grounding, agents explore the wrong paths. With it, they start with the map. Graph in Fabric is also now generally available, built as a relationship-first data modeling engine natively on OneLake.  It runs on GQL (ISO/IEC 39075), supports billions of relationships with sharded scale-out processing, and includes natural-language-to-GQL for non-technical users.  Paired with Ontology, Graph is the execution layer that lets AI agents reason consistently across connected business domains, from supply chain and fraud to customer intelligence and cybersecurity. The connective tissue extends further: Ontology MCP support lets external agents and tools connect through the Model Context Protocol. Foundry IQ agents can use the ontology as a governed knowledge source. Copilot Studio can call it as a tool.  Define business context once in Fabric IQ, reuse it across the full spectrum of agent experiences. By Yitzhak Kesselman and Chafia Aouissi Fabric IQ: The shared context layer for AI agents and real-time applications community.fabric.microsoft.c… Fabric IQ: The semantic layer powering trusted AI agents at enterprise scale community.fabric.microsoft.c… #FabricIQ #KnowledgeGraph #SemanticLayer #Ontology #GraphData -- Connected Data London 2026 | 11–12 November | Leonardo Royal Hotel London Tower Bridge 🎤 Share your work with the world's most passionate data community. The Call for Submissions is open.  connected-data.london/2026-c… 🎟 Tickets on sale now. Early bird discounts up to 30%. 2026.connected-data.london?u… 📺 Sponsorship opportunities available. Contact info@connected-data.london for details. #KnowledgeGraph #GraphRAG #Ontology #Graph #AI #DataScience #GraphDB #SemTech

Meet the new face of GFQL: The mythical kraken! Its many arms represent how graphs help us reach across many things, and special to GFQL’s GPU approach, the power of the kraken is legendary. Our young mascot is a small addition, but one we hope makes the project feel a little more recognizable and fun as our community keeps growing. At the same time, we’re moving the GFQL community over to the Graphistry Discord. Slack was great for getting started, but to solve limitations such as in searchable chat histories and easy invitation links, Discord gives us a better place to share ideas, help each other, show projects, and talk graph compute in real time. 👉 Join the Graphistry Discord: discord.gg/QXaFt8wk8Q Inside: • GFQL discussions & support • Graphistry Louie updates • Community demos & experiments • A place to help shape what comes next If you’re working with Cypher, GPU analytics, graph ML, cuGraph, Neo4j, Arrow, RAPIDS, or just exploring graph-shaped data in Python, come hang out. Happy graphing. #GFQL #Graphistry #GraphData #GraphComputing #OpenSource

中医知识图谱 v2 正式发布！🎉 相较 v1，本次在数据建模、实体关系设计与语义表达上全面优化，增强结构一致性与可扩展性；同时丰富中药、方剂与功效关联，提升查询与探索体验。 🔗 github.com/yasenstar/Chinese… 欢迎 Star ⭐、交流与共建！ #知识图谱 #中医 #GraphData #TraditionalMedicine #Medicine

Evolution of an EA MetaModel. 🏗️ I'm documenting the step-by-step build using @arrows_app and #Neo4j. The latest update (022) bridges the gap between Business Conceptual and Business Logical designs. See the progress on GitHub: 👉 github.com/yasenstar/EA_Meta… #ArchiMate #GraphData

🚀 We’re excited to share that our paper “Efficient Graph Data Access for Out-of-Memory GPU Streaming Graph Processing”, in collaboration with NUS & ByteDance, was selected as Best of VLDB 2025! 🎉 Congrats to Qiange Wang and Yongze Yan! 🙌 #VLDB2025 #GraphData

GraphRAG often fails due to pipeline sprawl, not retrieval. Atomic GraphRAG: run the pipeline as a single Cypher query in the DB, with 10x less orchestration code. Explainer by @mbudiselicbuda 👉 memgraph.com/blog/atomic-gra… #GraphData #Memgraph

Need vector search that scales without doubling storage? In this blog, David Iveković breaks down how Memgraph avoids duplicate vector copies and improves memory efficiency. Full write-up 👉 memgraph.com/blog/single-sto… #Memgraph #VectorSearch #VectorIndex #GraphData #GraphRAG #RAG

Workshop proposal accepted at #VLDB2026 #VLDB26: Agent Graph: 3rd International Workshop on Data Management Opportunities in Bringing Agents with Graph Data 2026 Workshop. Call for Papers coming soon #GraphData #LLMs #Agents (co-organizing with Yixiang Fan, Tianxing Wu, Da Yan).

안녕하세요 저소득층 방귀대장 뿌뿌웅뿌웅입니다. 진짜 2025년의 마지막 글 @graphprotocol 와 함께합니다! 새해를 맞아 굉장히 기대되는 Horizon의 업그레이드가 진행 됬다고 하더라고요! 같이 살펴 보면서 제 의견을 5가지의 갈래로 말해보는 시간을 가져보려 합니다! 1. 왜 Horizon이 필요했나: 서브그래프만으로는 부족해진 데이터 수요 The Graph는 2020년 런칭 이후 Subgraph 기반으로 대규모 쿼리를 처리하며 블록체인 데이터 접근 표준에 가까운 포지션을 만들었습니다! 최근 데이터 수요는 실시간 스트림·대규모 분석·커스텀 API·특수 쿼리 엔진처럼 더 다양한 종류가 생겼쥬! 그래서 문서도 이제 Subgraphs만으로 커버되지 않는 유즈케이스가 늘었다고 전제한다. 즉 Horizon은 서브그래프를 버리는 업그레이드가 아니라, 서브그래프 성공으로 증명된 경제,검증, 결제 메커니즘을 여러 데이터 서비스가 재사용하도록 확장하는 방향으로 잡았더라고요! 2.Horizon의 목표: 모든 블록체인 데이터 서비스를 위한 모듈형 프로토콜 Horizon은 Core staking,Core payments, Data Service Framework라는 재사용 가능한 프로토콜 프리미티브를 분리해, 서브그래프가 아니라도 임의의 데이터 서비스가 동일한 보안·결제 층을 이용하게 합니다! 정리하면: Subgraph는 계속 운영되지만, The Graph의 프로토콜 본체는 이제 멀티 서비스형으로 재정의되는 것이죠! 3. 핵심 구조 3가지: Core Staking, Unified Payments, Data Service Framework가 중요한데요! Core Staking Protocol: 경제적 보안의 기준을 서브그래프 인덱싱”에만 묶지 않고, 데이터 제공자와 소비자 간 데이터 교환 전반을 보호하는 공통 스테이킹 층으로 확장하는 시스템이더라고요! Core Payments Protocol(통합 결제): 서비스별로 결제가 쪼개지지 않게, 여러 데이터 서비스가 같은 결제와 정산 레일을 공유합니다! The GraphData Service Framework: 누구나 새 데이터 서비스를 만들 때, 보안·결제 같은 기반을 다시 만들지 않고, 프레임워크 위에서 조립하듯 붙일 수 있게 하는 시스템입니다! 4. 서비스 스펙트럼 확장: Horizon은 서브그래프는 그대로, 하지만 프로토콜은 실시간 스트림·프리인덱스 API·애널리틱스·검증 서비스 등도 지원하고 있습니다! 실시간 이벤트는 Substreams과 Firehose 지갑과 마켓의 잔고·전송은 Token API 에서 진행하고 있죠! 5. 참여자·토크노믹스 관점 변화: GRT의 역할이 서브그래프에서 데이터 서비스 전반으로 Horizon은 서비스가 늘면 프로토콜을 통과하는 GRT fee가 늘 수 있는 가능성이 생기고 파라미터에 따라 burn이 발생할 수 있으며 거버넌스로 서비스별 발행 그리고 인센티브 배분도 가능하다고 합니다! 이 말은 곧 Indexer와 Delegator 같은 네트워크 참여자가 단일 수익원(서브그래프 쿼리)에만 의존하지 않고, 장기적으로는 여러 서비스의 fee 흐름에 노출될 수 있다는 설계 방향을 뜻합니다!! 2026년이 코앞입니다! 다들 새해 복 많이 받으세요! Happy new year!!

"Your output can be our input." - Weimo Liu, CEO @ PuppyGraph Graph analytics 𝘀𝗵𝗼𝘂𝗹𝗱𝗻’𝘁 require a whole second data pipeline 😅 But the usual setup is still: data lake ➝ ETL ➝ graph database. PuppyGraph skips all that 😎⏭️ PuppyGraph is a 𝗴𝗿𝗮𝗽𝗵 𝗾𝘂𝗲𝗿𝘆 𝗲𝗻𝗴𝗶𝗻𝗲, not a graph database. This means it sits 𝗱𝗶𝗿𝗲𝗰𝘁𝗹𝘆 atop your relational data as the graph layer, so you can: 🐾 Plug graph queries into your existing data stores, zero ETL required 💬 Query in real time as data streams into your tables 🚀 Stay fast at petabyte scale, with optimizations built for big data 🧠 Leverage the same data copy for both SQL analytics and multi-hop graph questions Check out more of Weimo’s insights from the most recent #ApacheIceberg meetup ❄️🎤 #ApacheIceberg #Lakehouse #DataEngineering #GraphAnalytics #GraphData #OpenTableFormats #ZeroETL #DataArchitecture

Geo-RDF Framework for Representing the Spatial Information of Bangladesh aircconline.com/ijwest/V16N4… airccse.org/journal/ijwest/v… #semanticweb #ontology #RDF #spatialdata #geoSPARQL #LinkedData #KnowledgeGraph #OWL #SPARQL #GraphData #GraphDB #GraphTechnology #LinkedOpenData #Ontology

Need better accuracy to run natural-language Q&A over multi-layered #GraphData? Skip the old "#LLM → #Cypher" chain. Go for an #MCP-based RAG app. Full walkthrough 👉 memgraph.com/webinars/lesson… Key highlights 👉 memgraph.com/blog/building-m… #Memgraph

Your friend’s friend’s friend knows you better than you think. @neo4j CTO Philip Rathle (@prathle) breaks down the wild research showing how deeply our networks influence us — sometimes more than our own facts. 🎥 Watch the full interview: youtu.be/rSZWIPtOcD4 #AI #Neo4j #GraphData #FounderCoHo #ConnectedWorld #BehaviorScience #Networks

DeepChem Equivariant: SE(3)-Equivariant Support in an Open-Source Molecular Machine Learning Library 1. The integration of SE(3)-equivariant neural networks into the DeepChem library represents a significant leap forward in molecular machine learning. These networks ensure that model outputs transform predictably with input coordinate changes, crucial for tasks like molecular property prediction and protein structure modeling. 2. The core innovation lies in the implementation of SE(3)-equivariant models such as SE(3)-Transformer and Tensor Field Networks within DeepChem. This modular infrastructure broadens the application of equivariant methods, making them accessible to scientists with minimal deep learning background. 3. The framework efficiently computes equivariant features using spherical harmonics and irreducible representations. This approach not only respects 3D geometric symmetries but also offers a practical and well-documented suite of tools for equivariant molecular modeling. 4. The implementation includes complete training pipelines and a toolkit of equivariant utilities, supported with comprehensive tests and documentation. This ensures robustness and fosters contributions from the community, making it easier to build, train, and evaluate models. 5. Experiments on the QM9 dataset demonstrate that DeepChem's SE(3)-equivariant models achieve comparable performance to state-of-the-art methods, highlighting the effectiveness of incorporating attention mechanisms into roto-translation-equivariant models. 6. Future work includes implementing LieConv in DeepChem-Equivariant to improve efficiency, particularly in basis construction. Caching precomputed bases in DeepChem's GraphData class is also proposed to avoid bottlenecks during training and improve scalability. 📜Paper: arxiv.org/abs/2510.16897 #DeepChem #EquivariantNeuralNetworks #MolecularMachineLearning #SE3Equivariance #OpenSource #MolecularModeling #DrugDiscovery #MaterialsScience

We’re live at the Neo4j Graph Summit London 2025. In County Hall, ETC Venue for a full day of graph technology, connected data insights, and real-world applications. If you’re here too – let’s connect! #Neo4j #GraphSummit #GraphData #ConnectedData #LondonEvents #WeAreLive

2. Excited to share two of our papers at #EMNLP25 #EMNLP2025: 2. "CoT-RAG: Integrating Chain of Thought and Retrieval-Augmented Generation to Enhance Reasoning in Large Language Models". Accepted at #EMNLP25Findings arxiv.org/abs/2504.13534 #LLMs, #KGs, #GraphData #CoT #RAG

1. Excited to share two of our papers at #EMNLP25 #EMNLP2025: 1. "Large Language Models Meet Knowledge Graphs for Question Answering: Synthesis and Opportunities" - (led by @DylanMa2018). Accepted at #EMNLP25Main arxiv.org/abs/2504.13534 #LLMs, #KGs, #GraphData #QA

- Turned panels like properties, preview3d, GraphData, Compile, into nodes with hotkeys to add/remove them when and where needed. (may change some of them later to static ones again)

I love (HATE) this specific bug I cannot cast the graph param to GraphData because the GraphData type isn't a tool(editor running) type

We’re heading to the Neo4j Graph Summit in London, County Hall. A full day dedicated to exploring the power of graph technology and the future of connected data. If you are there, let’s connect! #Neo4j #GraphSummit #GraphData